CN108563638A - A kind of microblog emotional analysis method based on topic identification and integrated study - Google Patents
A kind of microblog emotional analysis method based on topic identification and integrated study Download PDFInfo
- Publication number
- CN108563638A CN108563638A CN201810333907.8A CN201810333907A CN108563638A CN 108563638 A CN108563638 A CN 108563638A CN 201810333907 A CN201810333907 A CN 201810333907A CN 108563638 A CN108563638 A CN 108563638A
- Authority
- CN
- China
- Prior art keywords
- text
- microblogging
- word
- topic
- emotion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Abstract
The microblog emotional analysis method based on topic identification and integrated study that the invention discloses a kind of, this approach includes the following steps：It collects microblog data and is manually marked；Microblogging text is pre-processed by text data processing method；Optimal text subject number is selected by LDA topic relativity indexs and excavates text subject with LDA；Theme feature, affective characteristics and the sentence features analyzed for microblog emotional in conjunction with sentiment dictionary structure；Using features described above as the input feature vector variable of training AdaBoost algorithms grader is analyzed to establish microblog emotional.The method of the present invention has excavated microblogging text semantic information by deep, effectively increases text emotion nicety of grading.
Description
Technical field
The present invention relates to natural language processing technique more particularly to a kind of microblogging feelings based on topic identification and integrated study
Feel analysis method.
Background technology
Social media is fast-developing in recent years, more and more network users' selections social network-i i-platform such as microblogging,
Forum, shopping website etc. express individual opinion and Sentiment orientation.Microblogging becomes net because its spread speed is fast, social effectiveness is big
People's information propagates, the important channel of acquisition of information.For the public accident of some groups, netizen tends to express on microblogging
The view and opinion of oneself.Often the duration is long for this kind of event, and concern number is more, and huge, people are influenced in the network user
The mood expressed by network forms public opinion, may influence the development of event, in some instances it may even be possible to influence it is related personal or
The decision of tissue.The user of a large amount of fragment types of these in microblogging, which generates information, can reflect the evolutionary process and public sentiment of event
Fluctuation situation, the discussion topic of these accidents is tracked in microblogging, to microblogging comment analyze, can also original event
Evolution, real-time control netizen mood, reduce public contingent even to society negative effect.Therefore to microblogging text into
Row sentiment analysis can assist government to carry out network public-opinion monitoring, maintain social stability.
Sentence itself is all conceived to mostly to the research of Sentiment orientation analysis at present, can be described from excavations such as text, grammers
Feature in the feature of Sentiment orientation, such as common grammar property, sentence features, sentence.
In the above-mentioned methods, although having reached preferable emotional semantic classification effect, without excavating the Deep Semantics of text
Information.
Invention content
The technical problem to be solved in the present invention be for the defects in the prior art, provide it is a kind of based on topic identification and
The microblog emotional analysis method of integrated study.
The technical solution adopted by the present invention to solve the technical problems is：It is a kind of micro- based on topic identification and integrated study
Rich sentiment analysis method, includes the following steps：
1) it acquires microblogging text data from microblog and is pre-processed, obtain optimization content of text and optimization text
Content phrase；The microblogging text data includes microblogging body matter, microblogging comment content, micro- literary forwarding number and comment number；
The pretreatment includes the artificial mark commented on microblogging；It is described to be manually labeled as：To the feelings of every microblogging comment
Sense tendency carries out handmarking, if it is forward direction that this, which comments on Sentiment orientation, is labeled as 1, is otherwise labeled as 0；
2) by LDA (Latent Dirichlet Allocation) topic models to optimizing content of text in step 1)
And optimization content of text word carry out Modeling Calculation, identify microblogging text subject information, obtain LDA theme distributions probability and
LDA optimizes content of text word and theme distribution probability, is assessed according to the Semantic Similarity between high score word in each theme
Theme quality determines the subject categories of appropriate number, using every affiliated subject categories of microblogging text as microblog users emotion point
The theme feature of analysis；
3) the positive emotion word that the comment of every microblogging occurs, negative sense emotion word, adversative and no are extracted according to sentiment dictionary
Determine word, measure positive emotion word, the quantity of negative sense emotion word, adversative and negative word, builds affective characteristics and sentence features, and
In conjunction with the theme feature that step 2) is extracted, multiple features vector combination of the structure for microblog emotional analysis；
4) by the input feature vector that the multiple features Vector Groups cooperation described in step 3) is AdaBoost models, select effect optimal
Feature Combination Design Sentiment orientation analyzes grader, and is trained according to the microblogging comment data manually marked described in step 1)
End user's emotion recognition grader is obtained, is applied to sentiment analysis and works.
By said program, Text Pretreatment further includes text participle, removes stop words and unrelated character mistake in the step 1)
Filter.
By said program, the evaluation index topic relativity that LDA topic identifications use in the step 2) is UMass theme
Correlation：
Wherein, coherence (V) is the theme Relevance scores, and score (vi, vj, ∈) is that UMass modules calculate
The method of score, V indicate describe some theme set of words, ∈ is a smoothing factor, for ensure return score be
One real number；D (vi, vj) indicates that the microblogging text quantity for including word vi and vj, D (vj) are indicating the microblogging comprising word vj just
Literary quantity.
By said program, the sentiment dictionary be according to Hownet HowNet sentiment dictionaries and converged network prevalence vocabulary, it is whole
Four text documents of reason, including positive emotion word, negative sense emotion word, negative word, adversative.
By said program, multiple features vector is combined as in the step 3)：
featurei={ topici,emotioni,sentencei, tendencyi}(1≤i≤M)；
Wherein, M is that this microblogging comments on item number, and i is that microblogging comments on serial number, topiciBe the theme feature, emotioniFor feelings
Feel feature, sentenceiFor sentence features, tendencyiFor this microblog text affective tendency manually marked；
Wherein,
emotioni={ n_posi,n_negi}(1≤i≤M)
Wherein, n_posiThe quantity of positive emotion word, n_neg in being commented on for thisiFor the quantity of negative sense emotion word；
sentencei={ n_denyi,n_trai}(1≤i≤M)
Wherein, n_denyiThe quantity of negative word, n_tra in being commented on for thisiFor adversative quantity.
The beneficial effect comprise that：
The present invention is based on the microblog emotional analysis methods of topic identification and integrated study can deep enough excavation microblogging text language
Adopted information, with LDA Model Identification microblogging themes, using its affective characteristics and sentence features variable with definition as integrated study side
The input variable of method AdaBoost carries out classification based training, and the present invention obtains higher accuracy rate in Sentiment orientation identification.
Description of the drawings
Present invention will be further explained below with reference to the attached drawings and examples, in attached drawing：
Fig. 1 is the method flow diagram of the embodiment of the present invention.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, is not used to limit
The fixed present invention.
As shown in Figure 1, the present invention provides a kind of microblog emotional analysis method of topic identification and integrated study, including it is following
Step：
Step 1, using reptile method from Sina weibo platform gathered data, the microblog data includes in microblogging text
Appearance, microblogging comment content, micro- literary forwarding number and comment number.Then it is pre-processed to crawling content, to obtain optimization text
This content and optimization content of text phrase, finally obtain 688 microblogging texts, 1426 microblogging comment datas；
Preferably, Text Pretreatment method described in step 1 includes the text participle of microblogging text and comment text, goes to stop
Word, the unrelated character of filtering and the artificial mark of microblogging comment；
It is manually labeled as described in step 1：
Handmarking is carried out to the Sentiment orientation of every microblogging comment, if it is forward direction that this, which comments on Sentiment orientation, is marked
It is 1, is otherwise labeled as 0；
Step 2, it is carried out by LDA topic models to optimizing content of text and optimization content of text word described in step 1
Modeling Calculation identifies microblogging text subject information, obtains LDA theme distributions probability and LDA optimization content of text words and master
Distribution probability is inscribed, theme quality is assessed according to the Semantic Similarity between high score word in each theme, by every microblogging text
Theme feature of the affiliated subject categories as microblog users sentiment analysis determines that topic relativity score is most when theme number is 18
Height finally selects the related commentary under wherein 6 themes to carry out emotion recognition；
Preferably, the evaluation index topic relativities of LDA topic identifications described in step 2 are UMass topic relativities：
Wherein, V is the set of words for describing some theme, and ∈ is a smoothing factor to ensure that the score returned is one
Real number；D (vi, vj) refers to the microblogging text quantity comprising word vi and vj, and D (vj) indicates the microblogging text number for including word vj
Amount.
Step 3, it improves sentiment dictionary, emotion word, adversative that every microblogging comment occurs, no is extracted according to sentiment dictionary
Determine word, measures positive emotion word, negative sense emotion word, adversative, negative word quantity, build affective characteristics and sentence features, and tie
Close the theme feature that step 2 is extracted, multiple features vector combination of the structure for microblog emotional analysis；
Preferably, sentiment dictionary described in step 3 is Hownet HowNet sentiment dictionaries and converged network prevalence vocabulary, arrange
For four text documents, including positive emotion word, negative sense emotion word, negative word, adversative；
Multiple features vector described in step 3 is combined as：
featurei={ topici,emotioni,sentencei}(1≤i≤M)
Wherein, M is that microblogging comments on item number, and i is that microblogging comments on serial number, topiciBe the theme feature, emotioniFor emotion spy
Sign, sentenceiFor sentence features.
Affective characteristics described in step 3 are：
emotioni={ n_posi,n_negi}(1≤i≤M)
Wherein, n_posiThe quantity of positive emotion word, n_neg in being commented on for thisiFor the quantity of negative sense emotion word；
Sentence features described in step 3 are：
sentencei={ n_denyi,n_trai}(1≤i≤M)
Wherein, n_denyiThe quantity of negative word, n_tra in being commented on for thisiFor adversative quantity；
Step 4, according to the affective characteristics described in step 3, theme feature conduct described in sentence features and step 2
The input feature vector of AdaBoost models, one kind of AdaBoost Ensemble classifier methods, base of the trade-off decision tree as AdaBoost
Learner, using the microblogging comment data after mark as T base learner of initial training collection training, according to the performance of base learner
Training sample distribution is adjusted, the sample of classification error increases its corresponding weight, reduces the weight of correct classification samples,
New sample distribution is obtained, giving the sample distribution for changing weights to sub-classification device is trained.Repeat, until base
The T values that device number reaches specified in advance are practised, T Weak Classifier is obtained, finally merge this T Weak Classifier by respective weights
(boost) get up, as the last grader for carrying out emotional semantic classification.And number is commented on according to the microblogging manually marked described in step 1
According to being trained to obtain end user's emotion recognition grader, it is applied to network user's Sentiment orientation and analyzes work.
Preferably, mode input described in step 4 is characterized as：
commenti={ topici,n_posi,n_negi,n_denyi,n_trai,tendencyi}(1≤i≤M)
Wherein, M is that microblogging comments on item number, and i is that microblogging comments on serial number, topiciAffiliated microblogging theme, n_ are commented on for this
posiFor the positive emotion word quantity in i-th comment, n_negiFor negative sense emotion word quantity, n_denyiFor negative word number in sentence
Amount, n_traiFor adversative quantity in sentence.tendencyiFor this text Sentiment orientation manually marked.Such as：" wish to safety
Back ", input feature vector is (1,3,0,0,0,1), and expression belongs to theme 1, and there are three positive emotion words, negative sense emotion word, no
Determine word, the quantity of adversative is all 0, and the Sentiment orientation of whole comment is forward direction；For another example：" so many themes in the U.S. are that do not have
Reason is that the film for that the abnormal dacnomania for liking killing by maltreatment is actually to be derived from life, fearful ", input feature vector for (5,
1,2,1,1,0) it, indicating to belong to theme 5, positive emotion word has 1, and negative sense emotion word has 2,1 negative word, 1 adversative,
The Sentiment orientation of whole comment is negative sense.The accuracy that last AdaBoost models are classified in test set reaches 85%.
Compared with prior art, the present invention is based on the microblog emotional analysis methods of topic identification and integrated study can be deep enough
Microblogging text semantic information is excavated, with LDA Model Identification microblogging themes, by the affective characteristics and sentence features variable of itself and definition
As integrated learning approach AdaBoost input variable carry out classification based training, the present invention Sentiment orientation identification on obtain compared with
High accuracy rate.
It should be understood that for those of ordinary skills, it can be modified or changed according to the above description,
And all these modifications and variations should all belong to the protection domain of appended claims of the present invention.
Claims (5)
1. a kind of microblog emotional analysis method based on topic identification and integrated study, which is characterized in that include the following steps：
1) it acquires microblogging text data from microblog and is pre-processed, obtain optimization content of text and optimization content of text
Phrase；The microblogging text data includes microblogging body matter, microblogging comment content, micro- literary forwarding number and comment number；
The pretreatment includes the artificial mark commented on microblogging；It is described to be manually labeled as：Incline to the emotion of every microblogging comment
To handmarking is carried out, if it is forward direction that this, which comments on Sentiment orientation, it is labeled as 1, is otherwise labeled as 0；
2) Modeling Calculation is carried out to optimization content of text in step 1) and optimization content of text word by LDA topic models,
It identifies microblogging text subject information, obtains LDA theme distributions probability and LDA optimization content of text words and theme distribution is general
Rate assesses theme quality according to the Semantic Similarity between high score word in each theme, determines the subject categories of appropriate number,
Using every affiliated subject categories of microblogging text as the theme feature of microblog users sentiment analysis；
3) positive emotion word, negative sense emotion word, adversative and the negative word that every microblogging comment occurs are extracted according to sentiment dictionary,
The positive emotion word of metering, the quantity of negative sense emotion word, adversative and negative word, build affective characteristics and sentence features, and combine
The theme feature of step 2) extraction, multiple features vector combination of the structure for microblog emotional analysis；
4) by the input feature vector that the multiple features Vector Groups cooperation described in step 3) is AdaBoost models, effect optimal characteristics are selected
Combination Design Sentiment orientation analyzes grader, and is trained to obtain according to the microblogging comment data manually marked described in step 1)
End user's emotion recognition grader is applied to sentiment analysis and works.
2. the microblog emotional analysis method according to claim 1 based on topic identification and integrated study, which is characterized in that
Text Pretreatment further includes text participle, removes stop words and unrelated character filtering in the step 1).
3. the microblog emotional analysis method according to claim 1 based on topic identification and integrated study, which is characterized in that
The evaluation index topic relativity that LDA topic identifications use in the step 2) is UMass topic relativity：
Wherein, coherence (V) is the theme Relevance scores, and score (vi, vj, ∈) is that UMass modules calculate score
Method, V indicate describe some theme set of words, ∈ is a smoothing factor, for ensure return score be one
Real number；D (vi, vj) indicates that the microblogging text quantity for including word vi and vj, D (vj) indicate the microblogging text number for including word vj
Amount.
4. the microblog emotional analysis method according to claim 1 based on topic identification and integrated study, which is characterized in that
Sentiment dictionary is according to Hownet HowNet sentiment dictionaries and converged network prevalence vocabulary, four texts of arrangement in the step 3)
Document, including positive emotion word, negative sense emotion word, negative word, adversative.
5. the microblog emotional analysis method according to claim 1 based on topic identification and integrated study, which is characterized in that
Multiple features vector is combined as in the step 3)：
featurei={ topici,emotioni,sentencei, tendencyi}(1≤i≤M)；
Wherein, M is that this microblogging comments on item number, and i is that microblogging comments on serial number, topiciBe the theme feature, emotioniFor emotion spy
Sign, sentenceiFor sentence features, tendencyiFor this microblog text affective tendency manually marked；
Wherein,
emotioni={ n_posi,n_negi}(1≤i≤M)
Wherein, n_posiThe quantity of positive emotion word, n_neg in being commented on for thisiFor the quantity of negative sense emotion word；
sentencei={ n_denyi,n_trai}(1≤i≤M)
Wherein, n_denyiThe quantity of negative word, n_tra in being commented on for thisiFor adversative quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810333907.8A CN108563638B (en) | 2018-04-13 | 2018-04-13 | Microblog emotion analysis method based on topic identification and integrated learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810333907.8A CN108563638B (en) | 2018-04-13 | 2018-04-13 | Microblog emotion analysis method based on topic identification and integrated learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108563638A true CN108563638A (en) | 2018-09-21 |
| CN108563638B CN108563638B (en) | 2021-08-10 |
Family
ID=63535041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810333907.8A Expired - Fee Related CN108563638B (en) | 2018-04-13 | 2018-04-13 | Microblog emotion analysis method based on topic identification and integrated learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108563638B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109284381A (en) * | 2018-09-27 | 2019-01-29 | 南通大学 | The aspect viewpoint of fusion emoticon library and topic model passes judgement on attitude method for digging |
| CN109684646A (en) * | 2019-01-15 | 2019-04-26 | 江苏大学 | A kind of microblog topic sentiment analysis method based on topic influence |
| CN109885826A (en) * | 2019-01-07 | 2019-06-14 | 平安科技（深圳）有限公司 | Text term vector acquisition methods, device, computer equipment and storage medium |
| CN110634050A (en) * | 2019-09-06 | 2019-12-31 | 北京无限光场科技有限公司 | Method, device, electronic equipment and storage medium for identifying house source type |
| CN111310476A (en) * | 2020-02-21 | 2020-06-19 | 山东大学 | Public opinion monitoring method and system using aspect-based emotion analysis method |
| CN111859074A (en) * | 2020-07-29 | 2020-10-30 | 东北大学 | Internet public opinion information source influence assessment method and system based on deep learning |
| CN112434164A (en) * | 2020-12-03 | 2021-03-02 | 西安交通大学 | Network public opinion analysis method and system considering topic discovery and emotion analysis |
| CN112765350A (en) * | 2021-01-15 | 2021-05-07 | 西华大学 | Microblog comment emotion classification method based on emoticons and text information |
| CN113127643A (en) * | 2021-05-11 | 2021-07-16 | 江南大学 | Deep learning rumor detection method integrating microblog themes and comments |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199857A (en) * | 2014-08-14 | 2014-12-10 | 西安交通大学 | Tax document hierarchical classification method based on multi-tag classification |
| CN106815369A (en) * | 2017-01-24 | 2017-06-09 | 中山大学 | A kind of file classification method based on Xgboost sorting algorithms |
| CN107908715A (en) * | 2017-11-10 | 2018-04-13 | 中国民航大学 | Microblog emotional polarity discriminating method based on Adaboost and grader Weighted Fusion |
-
2018
- 2018-04-13 CN CN201810333907.8A patent/CN108563638B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199857A (en) * | 2014-08-14 | 2014-12-10 | 西安交通大学 | Tax document hierarchical classification method based on multi-tag classification |
| CN106815369A (en) * | 2017-01-24 | 2017-06-09 | 中山大学 | A kind of file classification method based on Xgboost sorting algorithms |
| CN107908715A (en) * | 2017-11-10 | 2018-04-13 | 中国民航大学 | Microblog emotional polarity discriminating method based on Adaboost and grader Weighted Fusion |
Non-Patent Citations (4)
| Title |
|---|
| BASSAM AL-SALEMI ET AL.: "LDA-AdaBoost.MH Accelerated AdaBoost.MH based on latent Dirichlet allocation for text catergorization", 《JOURNAL OF INFORMATION SCIENCE》 * |
| FANGYU GAI ET AL.: "Enhance AdaBoost Algorithm by Integrating LDA Topic Model", 《INTERNATIONAL CONFERENCE ON DATA MINING AND BIG DATA》 * |
| KEITH STEVENS ET AL.: "Exploring Topic Coherence over many models and many topics", 《PROCEEDINGS OF THE 2012 JOINT CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING AND COMPUTATIONAL NATURAL LANGUAGE LEARNING》 * |
| 李杉: "面向中文微博文本的情感极性判别方法研究", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109284381A (en) * | 2018-09-27 | 2019-01-29 | 南通大学 | The aspect viewpoint of fusion emoticon library and topic model passes judgement on attitude method for digging |
| CN109284381B (en) * | 2018-09-27 | 2023-12-08 | 南通大学 | Aspect perspective appreciative and detractive attitude mining method integrating expression symbol library and theme model |
| CN109885826A (en) * | 2019-01-07 | 2019-06-14 | 平安科技（深圳）有限公司 | Text term vector acquisition methods, device, computer equipment and storage medium |
| CN109684646A (en) * | 2019-01-15 | 2019-04-26 | 江苏大学 | A kind of microblog topic sentiment analysis method based on topic influence |
| CN110634050B (en) * | 2019-09-06 | 2023-04-07 | 北京无限光场科技有限公司 | Method, device, electronic equipment and storage medium for identifying house source type |
| CN110634050A (en) * | 2019-09-06 | 2019-12-31 | 北京无限光场科技有限公司 | Method, device, electronic equipment and storage medium for identifying house source type |
| CN111310476A (en) * | 2020-02-21 | 2020-06-19 | 山东大学 | Public opinion monitoring method and system using aspect-based emotion analysis method |
| CN111859074A (en) * | 2020-07-29 | 2020-10-30 | 东北大学 | Internet public opinion information source influence assessment method and system based on deep learning |
| CN111859074B (en) * | 2020-07-29 | 2023-12-29 | 东北大学 | Network public opinion information source influence evaluation method and system based on deep learning |
| CN112434164A (en) * | 2020-12-03 | 2021-03-02 | 西安交通大学 | Network public opinion analysis method and system considering topic discovery and emotion analysis |
| CN112434164B (en) * | 2020-12-03 | 2023-04-28 | 西安交通大学 | Network public opinion analysis method and system taking topic discovery and emotion analysis into consideration |
| CN112765350A (en) * | 2021-01-15 | 2021-05-07 | 西华大学 | Microblog comment emotion classification method based on emoticons and text information |
| CN113127643A (en) * | 2021-05-11 | 2021-07-16 | 江南大学 | Deep learning rumor detection method integrating microblog themes and comments |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108563638B (en) | 2021-08-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108563638A (en) | A kind of microblog emotional analysis method based on topic identification and integrated study | |
| CN107092596B (en) | Text emotion analysis method based on attention CNNs and CCR | |
| Ghosh et al. | Fracking sarcasm using neural network | |
| CN107330011A (en) | The recognition methods of the name entity of many strategy fusions and device | |
| CN107025299B (en) | A kind of financial public sentiment cognitive method based on weighting LDA topic models | |
| CN103207913B (en) | The acquisition methods of commercial fine granularity semantic relation and system | |
| CN109376251A (en) | A kind of microblogging Chinese sentiment dictionary construction method based on term vector learning model | |
| TW201737118A (en) | Method and device for webpage text classification, method and device for webpage text recognition | |
| CN106919673A (en) | Text mood analysis system based on deep learning | |
| CN109933664A (en) | A kind of fine granularity mood analysis improved method based on emotion word insertion | |
| CN110609983B (en) | Structured decomposition method for policy file | |
| CN109670041A (en) | A kind of band based on binary channels text convolutional neural networks is made an uproar illegal short text recognition methods | |
| CN107122349A (en) | A kind of feature word of text extracting method based on word2vec LDA models | |
| CN106096664A (en) | A kind of sentiment analysis method based on social network data | |
| CN108647225A (en) | A kind of electric business grey black production public sentiment automatic mining method and system | |
| CN110532563A (en) | The detection method and device of crucial paragraph in text | |
| CN109886270A (en) | A kind of case element recognition methods towards electronics folder notes text | |
| Bartle et al. | Gender classification with deep learning | |
| CN110750648A (en) | Text emotion classification method based on deep learning and feature fusion | |
| CN110134934A (en) | Text emotion analysis method and device | |
| CN107463703A (en) | English social media account number classification method based on information gain | |
| CN112434164B (en) | Network public opinion analysis method and system taking topic discovery and emotion analysis into consideration | |
| CN108733675A (en) | Affective Evaluation method and device based on great amount of samples data | |
| CN111339772B (en) | Russian text emotion analysis method, electronic device and storage medium | |
| Forsyth | Automatic readability prediction for modern standard Arabic |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210810 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |